2010: PHP, CSS, Vanilla JS, and a LAMP Server.

Ah, the simple life.

2016: Node.js, React, Vue, Angular, AngularJS, Polymer, Sass, Less, Gulp, Bower, Grunt.

I can't handle this, I'm shifting domains to Machine Learning.

2017: Numpy, Scipy, TensorFlow, Theano, Keras, Torch, CNNs, RNNs, GANs and LOTS AND LOTS OF MATH!

Okay, okay. Calm down there fella.
JavaScript doesn't seem that complicated now, does it? 🙈

Finally program running perfectly 👍
But wait.... Lets add another feature...
New feature needs updated matplotlib...
Update matplotlib😒
Lost support with numpy😫🤔
Updated numpy 😫
Run program again...
Core dumped (segmentation fault)😶😶
Time to leave this planet

OK heavy rant on 'modern' software development coming! --> don't take it to seriously though :-)

Electron... why does that shit exist? It is like stacking all the worst technologies available to mankind into an enormous pile of crap and polishing that turd to look like something wonderful. It is big, slow and overall AWFUL!

An example? ... Microsoft Teams :-( it burns your PC like fire and makes it squeal for mercy.

When a library/framework becomes the ultimate evolution of abstraction layer upon abstraction layer and it simply should stop to exist and a reset button needs to be pressed.

I would love to see some research on the real world environmental impact that all those shitty slow and bloated web technologies have.

Solution:

Software energy label!

C, C++ and Rust e.t.c. and all accompanying efficient UI libraries should be the only languages/implementations allowed to get a A, B and C label.

Python (without C libraries like Numpy), JavaScript and all those other slow interpreted scripting/Web API nonsense should get a D, E or F label by default.

Have fun!

Numpy and tensorflow eating my ram. I need to upgrade it. #8gigsnotenough

bro just learn C bro I promise it's all smooth sailing bro haha lol just take up HTML with CSS bro its a piece of cake bro what bro lol just start coding up differential equations with numpy library haha its so simple bro just start with Ruby bro it will take only couple days bro what lol bro take this aeronautical course on how to code an airplane simulation bro its so simple bro just start algorithms on cryptography bro its so easy i cant bro just start writing drivers for printers bro haha lol just start writing a bootloader for a new Linux distro bro lol haha easy bro just make a billion dollar company bro haha its so simple.
keep going bro haha invent your own JS framework over a billion existing ones haha bro typescript is so easy bro lol what u say take up redis bro go from the first command bro learn mongodb and mysql together bro its so simple.

but bro don't try to master JS bro .. u will regret it forever bro.

When you encounter a bug in your code while writing a test and you have absolutely no idea what's wrong...

...and then you see it's a type problem.

Saw this on a new dev's desk ... Yea, he's fired

What I have: [a, b, c]
What I need: [[a],[b],[c]]
No documentation
Just numpy things ¯\_(ツ)_/¯

after beginning to learn numpy , i believe these packages were really created by some clown of a circus xD.
Everything is sooooo entertaining!!!

i learned java 3 years ago, but today if i had to crap out some crazy java or c++ expert , i would tell him about numpy's arrays...
Like , "hey dude python has this cool data structure in the numpy library called arrays, which can hold any datatypes in a kind of arraylist like fashion, and you can convert them from 1 dimensional to 1000 dimensional in just 1 line , and also do you know we can select any column with just array[position]? and even this position does not needs to be an integer, you can use a list , like array[[1,2,3]] will give you elements at array[1],array[2],array[3], and...."
wait, why is my friend dead ? xD
HAhahahaha

Does anyone else here hate people who use numpy panda and tensorflow and call themselves data scientists ??
Cuz I hate 'em. There are so many researchers who work day and night to figure out the math and algos which go into these libraries. These researchers are real data scientists.
If computerss sciemce would have been a religion, then just using these stupid libraries and claiming you are a data scientist would be blasphemy.

Man....I keep up with this strange love hate relationship I have with Python....

Last night it was python that literally wrote my homework: define all possible equivalent partition tables with cause and effect analysis and boundary value checks for a program. The whole thing wrote itself and all I had to do was verify the inputs. Something that I was able to do using jupyter with pandas and numpy. On one hand, I despise the lack of static typing and use of whitespace as a block delimiter. On the other I cannot but help feeling a high level of gratitude over the language and its high availability and ease of use for this.

Sure, I could have used other tools, but this language has dominated hardcore in this regard enough to the point of not considering it being a crime against humanity.

I know I’m not the only Numpy lover in the house. It deserves its own song…I’m pretending I haven’t already written one…

How deep does the rabbit hole go?

Problem: Convert numpy array containing an audio time series to a .wav file and save on disk

Error 1:
Me: pip install "stupid package"
Console: Can't pip, behind a proxy

Me: Finds workaround after several minutes

Error 2:
Conversion works, but audio file on disk doesn't work
Encoding Error only works with array of ints not floats
BUT I NEED IT TO BE FLOATS

Looks for another library
scikits.audiolab <- should work
Me: pip --proxy=myproxy:port install "this shit"
Command Line *spits back huge error*

Googles error <- You need to install this package with a .whl file
Me: Downloads .whl file <- pip install "filename".whl
Command Line: ERROR: scikits.audiolab-0.11.0-cp27-cp27m-win32.whl is not a supported wheel on this platform.

Googles Error <- Need to see supported file formats
Me: python -c "import pip; print(pip.pep425tags.get_supported())"
Console: AttributeError: module 'pip' has no attribute 'pep425tags'

Googles Error <- Use another command for pip v10
Me: python -c "import pip._internal; print(pip._internal.pep425tags.get_supported())"
Console: complies

Me: pip install "filename".whl
Console: complies

Me: *spends 30 minutes to find directory where I should paste .dll file*
Finds Directory (was hidden btw), pastes file

Me: Runs .py file
Console: from version import version as _version ModuleNotFoundError: No module named 'version'

Googles Error <- Fix is: "just comment out the import statement"
Me: HAHAHAHAHAHA
Console: HAHAHAHAHA

Unfortunately this shit still didn't work after two hours of debugging, lmao fuck this

I love python, but I hate dealing with python dependencies, especially on Windows.
I was tinkering and researching with neural networks, so I wanted to try out pybrain. I wrote my project, with pybrain installed via pip, and tried to build it.
Oh, what's that? Pybrain doesn't work with python 3? Well I'll download the version that's supposed to. Oh, that version has a deprecated numpy api? Let me just install those other resources. Oh, that requires a broken module that has no publicly available source?
Let's try python 2. Oh, now that's working, I just need to export environment variables for some "bls source". Some quick Google searching and the only solution that would work is building a bunch of cywgin modules by hand. That's fine, I have an ubuntu partition.
An hour later I'm compiling FORTRAN dependencies on Ubuntu.
Coding time: 1 hour
Dependency time: 3 hours

fuck off with the “do x in y lines of python code” it’s getting so goddamn annoying. yes python is concise. yes libraries do everything. you don’t need to show off someone else’s work with clickbait.

everything is like

“make a web server in 2 lines of python code”

import http.server
server = http.server.serve()

“mine bitcoin in 2 lines of python code”
import bitcoinminer
bitcoinminer.mine()

“do crazy math with 4 lines of python code”
import complex
import numpy
num1 = 1
num2 = 1
num3 = complex.addVectorMagnitudes(num1, num2)

Sorry Google, you got it wrong this time ....

Oh my gosh, look at that function definition ...
Oh my gosh, look at that variable ...
Oh my gosh, look at that zone ...
Oh my gosh, look at that long ...
Oh my gosh, look at that short ...
Oh my gosh, look at that stop ... is more my style.

python machine learning tutorials:
- import preprocessed dataset in perfect format specially crafted to match the model instead of reading from file like an actual real life would work
- use images data for recurrent neural network and see no problem
- use Conv1D for 2d input data like images
- use two letter variable names that only tutorial creator knows what they mean.
- do 10 data transformation in 1 line with no explanation of what is going on
- just enter these magic words
- okey guys thanks for watching make sure to hit that subscribe button

ehh, the machine learning ecosystem is burning pile of shit let me give you some examples:

- thanks to years of object oriented programming research and most wonderful abstractions we have "loss.backward()" which have no apparent connection to model but it affects the model, good to know
- cannot install the python packages because python must be >= 3.9 and at the same time < 3.9
- runtime error with bullshit cryptic message
- python having no data types but pytorch forces you to specify float32
- lets throw away the module name of a function with these simple tricks:
"import torch.nn.functional as F"
"import torch_geometric.transforms as T"
- tensor.detach().cpu().numpy() ???
- class NeuralNetwork(torch.nn.Module):
def __init__(self):
super(NeuralNetwork, self).__init__() ????
- lets call a function that switches on the tracking of math operations on tensors "model.train()" instead of something more indicative of the function actual effect like "model.set_mode_to_train()"
- what the fuck is ".iloc" ?
- solving environment -/- brings back memories when you could make a breakfast while the computer was turning on
- hey lets choose the slowest, most sloppy and inconsistent language ever created for high performance computing task called "data sCieNcE". but.. but. you can use numpy! I DONT GIVE A SHIT about numpy why don't you motherfuckers create a language that is inherently performant instead of calling some convoluted c++ library that requires 10s of dependencies? Why don't you create a package management system that works without me having to try random bullshit for 3 hours???
- lets set as industry standard a jupyter notebook which is not git compatible and have either 2 second latency of tab completion, no tab completion, no documentation on hover or useless documentation on hover, no way to easily redo the changes, no autosave, no error highlighting and possibility to use variable defined in a cell below in the cell above it
- lets use inconsistent variable names like "read_csv" and "isfile"
- lets pass a boolean variable as a string "true"
- lets contribute to tech enabled authoritarianism and create a face recognition and object detection models that china uses to destroy uyghur minority
- lets create a license plate computer vision system that will help government surveillance everyone, guys what a great idea

I don't want to deal with this bullshit language, bullshit ecosystem and bullshit unethical tech anymore.

I told interns in my startup to code a GAN only using Numpy. I received 4 resignation letters the next day

Okay! Got my numpy pdf, theano pdf and my theano deep learning pdf! It’s time to get reading for 1111111111111111111111111111111111111 hours. Wow! I’m really getting deep into “deep learning” learning! Ok, I’ll quit now...

When I first began with Python I really missed the static typed checking from Java, I barely know anything about a returned object from a method and have to read the API extensively for every new library.

After a while I finally understand why Python is so powerful, the combination of dynamic typed language and rich default methods make the language unbeatable for your productivity.

While Java's Object only has toString(), hashCode(), equals() or clone(), Python's basic Class has every fucking method for every scenario I could ever image. No wonder that libraries like numpy or pandas work so well and fluidly.

Reading through one of my posts I’ve realized how much ego programmers can actually have. Guys, some of you have already mastered or grasped more than just the foundations of the industry standard languages, as well as developed a very solid intuition behind some design patterns and a solid understanding of some frameworks and libraries, say NumPy, say React... we get it.

You don’t have to be such condescending assholes and be offended by some of the jokes we, programming beginners, make to release stress or just to have fun.

You already have some amazing developer and engineering skills. Do not ruin it with such a detrimental attitude; I make this post because I myself have made this mistake, and I still do to this day. But if what I’ve felt reading your comments is what non-programming people feel when around me, I wouldn’t be surprised if I found that some people hated me or just wanted to kill me.

I don’t know if this will get downvot’d or if more people think like this. But I needed to share this, even just as a reflection of my very own attitude.

Thank you for your time,

D.

Stuck at dealing with a huge amount of images again. 🤦
No idea how quickly I can get this object classification nn up and running, as it seems I have forgotten how to do shit. 🙄😒

Me using numpy.polyfit() to fit a function to some data. Blame polyfit for not being accurate, search the web to find any related problems, as the polynom doesn't fit at all. Polynom is of 2nd degree, polyfit becomes unstable at about degree 20?? Try Polynomial.fit() ... same results. How can they all be wrong???See a little typo in my code, calculating the polynom points. Fix it, everything works.... wasted 2 and a half weeks because of this error.🤦🤦

Y'all can bash me for it, but Python is one language that ought to be banned along with Javascript...

Amount of times that it breaks or have incomplete implementation is absurd. I just had to deal with idiotic developer who just love to break backward compatibility (looking at you numpy), by changing the type or function name by literally one letter which break older software written in Python that were still in use. (They never specify version for dependencies.) The best part is when they intentionally delete older dependency anyway even if the version is specified.

There's a reason why I do things in C language rather than any other languages, one of the big thing about it is that almost every libraries/code have kept backward compatibility in mind.

Numpy saved me 😌😌, as always...

Whatever the f is wrong with numpy devs!!!!

Like seriously bro....

I can't import the effing sklearn.decomposition to do some basic PCA and the best solution out there is to downgrade it to version 1.16.1. Like hell!!!!

Issue has been known since last year, but guess who cares effukers

Just got accepted as a Tutor. I have to teach PhD students in medical field SciKit package for image processing. Been coding in Python using pandas and numpy for years, but I know jack shit on SciKit.

I applied just for fun and got the position. Now I am fucking terrified.

Meanwhile I rejected a Teaching Assistant position because of this one.

Ideas I've had over the years that could pan out and be useful:

SMS-DB: Stands for SMS-Data Burst. Used to allow those with low cell signal or no data plan to transfer data between a phone and some client via the standard SMS text space. Would be slow, but would act kinda like dial-up over SMS (as mobile lines are compressed on all service levels, even LTE, so traditional dial-up wouldn't work!) I have a general idea on how packets would be laid out, but that's about it so far...

everything2PNG: Allows one to transpose any file's data into a PNG with a 3 byte per pixel (full color RGB), which allows for a "compression" of sorts (about 91, 93% on preliminary tests) AND allowing further, more efficient compression of the resulting file. (Plus... it's just kinda cool to see files transposed as PNGs.) I actually have a simple transposer to go to PNG, but can't yet go back. Large files (around 600MB) use upwards of 4GB with efficient paging and other optimizations via NumPy so far, so it's not *viable* yet, but it's coming along nicely.

RPi-GPIO Interconnection Bus: A master/slave or round robin method to allow for Raspberry Pis to communicate using GPIO, which can help free up network bandwidth in RPi cloud computing clusters. At most, this'd allow for 4 bits used for pushing to the GPIO "bus", and 4 bits used for pulling from the "bus". 8 pins total are usually unused minimum, so either 3 or 4 pins for upload, 3 or 4 for download, and potentially 1 or 2 for commands, general non-data communication, etc. I made a version of this concept using Round Robin for a client, but it was horribly slow. (I also don't have distribution rights for the code, so i'm working from scratch.) Definitely doable.

I'm not a data scientist but lately I've learned NumPy, Pandas and now I'm learning Matplotlib and Seaborn and after years of Excel the improvement is astounding.

Excel is far easier to approach (I casually use it since I was 6) but once you need to do more advanced stuff it requires a lot of tricks and workarounds which needs to be memorized and are hard to find just by reasoning or are straight impossible without the use of macros which introduces many compatibility issues.

Pandas on the other hand is harder to approach but once you learn the concepts between its basic data structures you can do a lot with little "Google-Fu".

Fuck yeah ... I have uploaded my major computation file to S3 and create Lambdas from those files(includes numpy and pandas also) and now I have only routes and invoke strategies in my EC3 .. looking for cost reduction....

Right now, everything. I started at a Consulting firm because I expected many new problems to tackle, solutions to develop and generally to always have a fire burning underneath my ass but instead I always develop the same standard bullshit.

I miss the days in my old job when there was just a problem and the task to solve it. When I stared down giant amounts of data, just KNOWING that somewhere in that mess is some structure I could exploit and that short moment of inspiration when I finally pinpointed it. The rush of endorphins when the solution became clear and everything fell into place to form a beautiful pattern amidst the chaos test data, git commits and numpy arrays.

Now its just "Yeah, would you just write another selenium testsuite that throws out fail or pass and wastes all the information because the only reason I'm a testmanager is because I'm too incompetent to do anything else and not my passion for the field".

The constant, mind numbing repetition of always the same patterns where the occasional dynamic element that becomes stale is the highlight of my work week... I would have never thought that making good money with easy work would ever get me as close to depression as it did.

#get unique images ids
images_ids = np.unique(images_df.index)

Dear developer who wrote the code I'm looking at,
thanks, I really need comments like this one. I was wandering lost in 1500 lines of code, looking for an explaination of what the actual fuck the code is doing, and there I see you, comment. It's not like I want to know what the hundreds of lines functions do, who cares about that. What I needed to know, what shed light on this dark forest, is what the numpy functions do, because as you certainly know dear developer, such functions are really hard to comprehend, lacking of documentation.
Thanks.

So, I'm using Vpython for my physics class.
The good thing is that I love that. The bad thing is that the last update to vpython was in 2015.
Well, I update my system yesterday and apparently one of the libraries got an API update and that broke Vpython and can't be used again T.T
I'm trying to fix the code at the moment ;-;

I have started to work on pandas and numpy , should I work on scipy and sklearn or should make a strong fundamentals on numpy and pandas?

An open source technical graphing and analysis software using Matplotlib numpy and Scipy backends.

I love serverless functions but I'm so tired of complex orchestration, juggling event parameters and now scipy+numpy+pandas exceeds size limit of 250MB..

Feel like cramming it all in a monolith like the geezers of yore and be done with it

Who on earth decided, that float64 is a suitable default datatype for one-hot vectors in numpy?
That's what I deserve for relying on reasonable implicit behaviour

Python User-
C:\WINDOWS\system32>pip install scikit-image
CMD/Bash-
Collecting scikit-image
Downloading scikit-image(12.6MB)
██████████████████████████████
Collecting numpy
Downloading numpy(1.3MB)
██████████████████████████████
Collecting matplotlib
Downloading matplotlib(1.3KB)
██████████████████████████████
Collecting decorator
Downloading decorator(6.8MB)
██████████████████████████████
Collecting imageio
Downloading imageio(3.6MB)
██████████████████████████████
Collecting cycler
Downloading cycler(2.9MB)
██████████████████████████████

Installing collected packages cycler, imageio, decorator, matplotlib, numpy, scikit-image
Successfully installed cycler, imageio, decorator, matplotlib, numpy
Failed to load DLL of scikit-image.
C:\WINDOWS\system32>

Come on numpy
Let’s go compensate

how do i get the installed numpy version for tensorflow that pip autoinstalled to not interfere with opencv which requires numpy >=1.2.0 ?
which btw i just installed.
and why is tensorflow a piece of crap ? LOL

can you please help me with this.
I'm creating dataset of [Leet words][1].
This code is for generating [Leet words][1]. it is working fine with less number of strings but I've nearly 3,800+ strings and my pc is not capable to do so. I've Tried to run this on cloud(30gb RAM) not worked for me. but I think possible solution is to convert this code into numpy but I don't know how. if you know any other efficient way to do this it will be helpful.
Thanks!

from itertools import product
import pandas as pd
REPLACE = {'a': '@', 'i': '*', 'o': '*', 'u': '*', 'v': '*',
'l': '1', 'e': '*', 's': '$', 't': '7'}

def Leet2Combos(word):
possibles = []
for l in word.lower():
ll = REPLACE.get(l, l)
possibles.append( (l,) if ll == l else (l, ll) )
return [ ''.join(t) for t in product(*possibles) ]
s="""india
love
USA"""

words = s.split('\n')
print(words)
lst=[]
# ['india', 'love', 'USA']
for word in words:
lst.append(Leet2Combos(word))
k = pd.DataFrame(lst)
k.head()

I want to manipulate CSV files with Python and I was using NumPy, what I want to do is 3 columns, with an undetermined number of rows, and I want to be able to remove, add and edit every value, this is my questions:
Should I use NumPy? (if yes, please tell me how, I've been searching on google and I couldn't find anything of help! If not, please tell me what I should use,)

Can't seem to get the hang of numpy arrays..irritating the living HELL out of me rn

I am trying to decompose a 3D matrix using python library scikit-tensor. I managed to decompose my Tensor (with dimensions 100x50x5) into three matrices. My question is how can I compose the initial matrix again using the decomposed matrix produced with Tensor factorization? I want to check if the decomposition has any meaning. My code is the following:

import logging

from scipy.io.matlab import loadmat

from sktensor import dtensor, cp_als

import numpy as np

//Set logging to DEBUG to see CP-ALS information

logging.basicConfig(level=logging.DEBUG)

T = np.ones((400, 50))

T = dtensor(T)

P, fit, itr, exectimes = cp_als(T, 10, init='random')

// how can I re-compose the Matrix T? TA = np.dot(P.U[0], P.U[1].T)

I am using the canonical decomposition as provided from the scikit-tensor library function cp_als. Also what is the expected dimensionality of the decomposed matrices.